Research and development theory methods and models for forecasting time series with insurance increments

Abstract

The most actively developing area of the Russian economy is insurance. 213 insurance companies practicing in Russia have permission to provide insurance services for 2019. A clear-cut ascendancy in insurance premiums, which is proportional to inflation in the country, is an indicator of the increasing role of insurance in the market economy of the Russian Federation.

The article presents classic models for forecasting insurance time series, which obey normal laws, based on mathematical and instrumental methods, and on observations of the components of the predicted time series. The presented time series, possessing a long-term memory, is an exception to the rule.

The need to develop methods and models for forecasting time series of insurance with increments, as well as for their forecasting, is an urgent topic, since the lack of specific methods and models still creates problems for predicting incremental time series. In practice, classical forecasting methods give good results, thereby showing the necessity of using linear paradigm methods. As follows, the use of both a linear paradigm and a nonlinear paradigm is relevant, which gives us the opportunity to use them in a mixed form.